The effect of China's driving restrictions on air pollution: The role of a policy announcement without a stated expiration

Driving restrictions keep cars off the street largely on the basis of the last digit of the license plate number. This paper evaluates how different driving restriction policies affected the air quality of the 17 cities in Henan, one of China’s most populous provinces, from 2017 to 2019. I offer a novel way to categorize driving restrictions by making a distinction between cities that announced driving restrictions without a stated expiration and other cities that announced driving restrictions with a stated expiration. I provide some suggestive evidence on the exogeneity of timing, and using two-way fixed effects, I find significant heterogeneity in policy effectiveness. The policy reduced particulate matter concentration in cities that announced driving restrictions without a stated expiration, but had no effect in cities that announced driving restrictions with a stated expiration. To explain this difference, I build a model, which implies that driving restrictions announced without a stated expiration can induce people to pay a high fixed cost in exchange for a low transit time and not to choose non-compliance, two behavioral changes with the potential to further reduce air pollution. Thus, managing the duration of a policy and people’s expectations of its duration can matter crucially for the policy’s effectiveness.